Photon-starved snapshot holography

Yunping Zhang¹, Stanley H. Chan², Edmund Y. Lam¹

APL Photonics, May 2023
¹Department of Electrical and Electronic Engineering, The University of Hong Kong, Pok fu lam, Hong Kong SAR, China
²School of Electrical and Computer Engineering, Purdue University, West Lafayette, Indiana 47907, USA

Paper Code (coming soon)

Overview

See beyond the limits of conventional holographic imaging!

To push the capability of DH to extremely low photon flux situations, i.e., one photon per pixel or lower, we consider a new hologram formation mechanism with single-photon detection. It opens new possibilities for more advanced holography applications in situations such as probing live cells with minimal light interaction and high-speed volumetric tracking in flow cytometry, where the holograms suffer from poor pattern visibility. This work has been highlighted as a Scilight in the May 2023 issue of APL Photonics.

Optical system

For the first time, a quanta image sensor (QIS) is integrated into the traditional inline holographic setup as we used in our previous work [1], which is a newly developed image sensor with photon-number-resolving capability. It captures a stack of binary holographic frames created by limited photons with Poisson statistics. Among various photon-counting image sensors reported over the past decade, QIS stands out with its small pixel size (e.g. $\leq 1 μ m$ ), high photon detection efficiency, and ultra-low readout noise. For a comprehensive introduction to QIS, please refer to the review paper [2,3].

Physical setup

Physics-informed network

Integrating QIS in DH is not a trivial task, due to its binary signal detection mechanism and very different noise characteristics, which would require specifically designed reconstruction algorithms. In this work, we designed a physics-informed wavefront reconstruction network called PSHoloNet to reconstruct the object wavefront information from the QIS recordings. It is an unrolled network with task-oriented blocks and loss function. To demonstrate the feasibility through synthetic experiments, we show two DH applications, namely, 3D particle volumetric reconstruction and phase imaging. Please refer to the paper for more details.

Network Unrolling

Backbone Structure

Task-oriented blocks

Citation

@article{zhang2023photon,
  title={Photon-starved snapshot holography},
  author={Zhang, Yunping and Chan, Stanley H and Lam, Edmund Y},
  journal={APL Photonics},
  volume={8},
  number={5},
  pages={056106},
  year={2023},
  month = {May}
}

References

Yunping Zhang, Yanmin Zhu and Edmund Y. Lam,‘‘Holographic 3D particle reconstruction using a one-stage network’’,Applied Optics , vol. 61, no. 5, pp.B111-B120, November 2022.
Jiaju Ma, Stanley H. Chan, Eric R. Fossum, ‘‘Review of Quanta Image Sensors for Ultra-Low-Light Imaging’’, IEEE Journal of Electron Devices, vol. 69, no. 6, pp.2824-2839, June 2022.
Stanley H. Chan, ‘‘What Does a One-Bit Quanta Image Sensor Offer?’’, IEEE Trans. Computational Imaging, vol. 8, pp.770-783, Aug 2022.

Related Projects

You may also be interested in related projects from our group on holographic imaging:

Tianjiao Zeng, Yanmin Zhu, and Edmund Y. Lam, “Deep learning for digital holography: a review,” Optics Express, vol. 29, no. 24, pp. 40572–40593, November 2021.
Tianjiao Zeng, Hayden K.-H. So, and Edmund Y. Lam, “RedCap: residual encoder-decoder capsule network for holographic image reconstruction,” Optics Express, vol. 28, no. 4, pp. 4876–4887, February 2020.
Zhenbo Ren, Zhimin Xu, and Edmund Y. Lam, “End-to-end deep learning framework for digital holographic reconstruction,” Advanced Photonics, vol. 1, no. 1, pp. 016004(1–12), January 2019.